par
Hubert Roksor » 06 juil. 2007, 01:05
Tu peux réutiliser ces fonctions tirées de phpBB 3.0
/**
* Converts an NCR to a UTF-8 char
*
* @param int $cp UNICODE code point
* @return string UTF-8 char
*/
function utf8_chr($cp)
{
if ($cp > 0xFFFF)
{
return chr(0xF0 | ($cp >> 18)) . chr(0x80 | (($cp >> 12) & 0x3F)) . chr(0x80 | (($cp >> 6) & 0x3F)) . chr(0x80 | ($cp & 0x3F));
}
else if ($cp > 0x7FF)
{
return chr(0xE0 | ($cp >> 12)) . chr(0x80 | (($cp >> 6) & 0x3F)) . chr(0x80 | ($cp & 0x3F));
}
else if ($cp > 0x7F)
{
return chr(0xC0 | ($cp >> 6)) . chr(0x80 | ($cp & 0x3F));
}
else
{
return chr($cp);
}
}
/**
* Convert Numeric Character References to UTF-8 chars
*
* Notes:
* - we do not convert NCRs recursively, if you pass & it will return &
* - we DO NOT check for the existence of the Unicode characters, therefore an entity may be converted to an inexistent codepoint
*
* @param string $text String to convert, encoded in UTF-8 (no normal form required)
* @return string UTF-8 string where NCRs have been replaced with the actual chars
*/
function utf8_decode_ncr($text)
{
return preg_replace_callback('/&#([0-9]{1,6}|x[0-9A-F]{1,5});/i', 'utf8_decode_ncr_callback', $text);
}
/**
* Callback used in decode_ncr()
*
* Takes a NCR (in decimal or hexadecimal) and returns a UTF-8 char. Attention, $m is an array.
* It will ignore most of invalid NCRs, but not all!
*
* @param array $m 0-based numerically indexed array passed by preg_replace_callback()
* @return string UTF-8 char
*/
function utf8_decode_ncr_callback($m)
{
$cp = (strncasecmp($m[1], 'x', 1)) ? $m[1] : hexdec(substr($m[1], 1));
return utf8_chr($cp);
}
utf8_decode_ncr() remplacera tous les { par le character correspondant en UTF-8.
Tu peux réutiliser ces fonctions tirées de phpBB 3.0
[php]/**
* Converts an NCR to a UTF-8 char
*
* @param int $cp UNICODE code point
* @return string UTF-8 char
*/
function utf8_chr($cp)
{
if ($cp > 0xFFFF)
{
return chr(0xF0 | ($cp >> 18)) . chr(0x80 | (($cp >> 12) & 0x3F)) . chr(0x80 | (($cp >> 6) & 0x3F)) . chr(0x80 | ($cp & 0x3F));
}
else if ($cp > 0x7FF)
{
return chr(0xE0 | ($cp >> 12)) . chr(0x80 | (($cp >> 6) & 0x3F)) . chr(0x80 | ($cp & 0x3F));
}
else if ($cp > 0x7F)
{
return chr(0xC0 | ($cp >> 6)) . chr(0x80 | ($cp & 0x3F));
}
else
{
return chr($cp);
}
}
/**
* Convert Numeric Character References to UTF-8 chars
*
* Notes:
* - we do not convert NCRs recursively, if you pass & it will return &
* - we DO NOT check for the existence of the Unicode characters, therefore an entity may be converted to an inexistent codepoint
*
* @param string $text String to convert, encoded in UTF-8 (no normal form required)
* @return string UTF-8 string where NCRs have been replaced with the actual chars
*/
function utf8_decode_ncr($text)
{
return preg_replace_callback('/&#([0-9]{1,6}|x[0-9A-F]{1,5});/i', 'utf8_decode_ncr_callback', $text);
}
/**
* Callback used in decode_ncr()
*
* Takes a NCR (in decimal or hexadecimal) and returns a UTF-8 char. Attention, $m is an array.
* It will ignore most of invalid NCRs, but not all!
*
* @param array $m 0-based numerically indexed array passed by preg_replace_callback()
* @return string UTF-8 char
*/
function utf8_decode_ncr_callback($m)
{
$cp = (strncasecmp($m[1], 'x', 1)) ? $m[1] : hexdec(substr($m[1], 1));
return utf8_chr($cp);
}[/php]
utf8_decode_ncr() remplacera tous les { par le character correspondant en UTF-8.